Image feature selection by a genetic algorithm: application to classification of mass and normal breast tissue

Mammographic mass segmentation using fuzzy contours

BACKGROUND AND OBJECTIVE

Accurate mass segmentation in mammographic images is a critical requirement for computer-aided diagnosis systems since it allows accurate feature extraction and thus improves classification precision.

METHODS

In this paper, a novel automatic breast mass segmentation approach is presented. This approach consists of mainly three stages: contour initialization applied to a given region of interest; construction of fuzzy contours and estimation of fuzzy membership maps of different classes in the considered image; integration of these maps in the Chan-Vese model to get a fuzzy-energy based model that is used for final delineation of mass.

RESULTS

The proposed approach is evaluated using mass regions of interest extracted from the mini-MIAS database. The experimental results show that the proposed method achieves an average true positive rate of 91.12% with a precision of 88.08%.

CONCLUSIONS

The achieved results show high accuracy in breast mass segmentation when compared to manually annotated ground truth and to other methods from the literature.

Toward Understanding the Size Dependence of Shape Features for Predicting Spiculation in Lung Nodules for Computer-Aided Diagnosis

We analyze the importance of shape features for predicting spiculation ratings assigned by radiologists to lung nodules in computed tomography (CT) scans. Using the Lung Image Database Consortium (LIDC) data and classification models based on decision trees, we demonstrate that the importance of several shape features increases disproportionately relative to other image features with increasing size of the nodule. Our shaped-based classification results show an area under the receiver operating characteristic (ROC) curve of 0.65 when classifying spiculation for small nodules and an area of 0.91 for large nodules, resulting in a 26% difference in classification performance using shape features. An analysis of the results illustrates that this change in performance is driven by features that measure boundary complexity, which perform well for large nodules but perform relatively poorly and do no better than other features for small nodules. For large nodules, the roughness of the segmented boundary maps well to the semantic concept of spiculation. For small nodules, measuring directly the complexity of hard segmentations does not yield good results for predicting spiculation due to limits imposed by spatial resolution and the uncertainty in boundary location. Therefore, a wider range of features, including shape, texture, and intensity features, are needed to predict spiculation ratings for small nodules. A further implication is that the efficacy of shape features for a particular classifier used to create computer-aided diagnosis systems depends on the distribution of nodule sizes in the training and testing sets, which may not be consistent across different research studies.

The Applications of Genetic Algorithms in Medicine

A great wealth of information is hidden amid medical research data that in some cases cannot be easily analyzed, if at all, using classical statistical methods. Inspired by nature, metaheuristic algorithms have been developed to offer optimal or near-optimal solutions to complex data analysis and decision-making tasks in a reasonable time. Due to their powerful features, metaheuristic algorithms have frequently been used in other fields of sciences. In medicine, however, the use of these algorithms are not known by physicians who may well benefit by applying them to solve complex medical problems. Therefore, in this paper, we introduce the genetic algorithm and its applications in medicine. The use of the genetic algorithm has promising implications in various medical specialties including radiology, radiotherapy, oncology, pediatrics, cardiology, endocrinology, surgery, obstetrics and gynecology, pulmonology, infectious diseases, orthopedics, rehabilitation medicine, neurology, pharmacotherapy, and health care management. This review introduces the applications of the genetic algorithm in disease screening, diagnosis, treatment planning, pharmacovigilance, prognosis, and health care management, and enables physicians to envision possible applications of this metaheuristic method in their medical career.].

A new and fast image feature selection method for developing an optimal mammographic mass detection scheme

PURPOSE

Selecting optimal features from a large image feature pool remains a major challenge in developing computer-aided detection (CAD) schemes of medical images. The objective of this study is to investigate a new approach to significantly improve efficacy of image feature selection and classifier optimization in developing a CAD scheme of mammographic masses.

METHODS

An image dataset including 1600 regions of interest (ROIs) in which 800 are positive (depicting malignant masses) and 800 are negative (depicting CAD-generated false positive regions) was used in this study. After segmentation of each suspicious lesion by a multilayer topographic region growth algorithm, 271 features were computed in different feature categories including shape, texture, contrast, isodensity, spiculation, local topological features, as well as the features related to the presence and location of fat and calcifications. Besides computing features from the original images, the authors also computed new texture features from the dilated lesion segments. In order to select optimal features from this initial feature pool and build a highly performing classifier, the authors examined and compared four feature selection methods to optimize an artificial neural network (ANN) based classifier, namely: (1) Phased Searching with NEAT in a Time-Scaled Framework, (2) A sequential floating forward selection (SFFS) method, (3) A genetic algorithm (GA), and (4) A sequential forward selection (SFS) method. Performances of the four approaches were assessed using a tenfold cross validation method.

RESULTS

Among these four methods, SFFS has highest efficacy, which takes 3%-5% of computational time as compared to GA approach, and yields the highest performance level with the area under a receiver operating characteristic curve (AUC) = 0.864 ± 0.034. The results also demonstrated that except using GA, including the new texture features computed from the dilated mass segments improved the AUC results of the ANNs optimized using other three feature selection methods. In addition, among 271 features, the shape, local morphological features, fat and calcification based features were the most frequently selected features to build ANNs.

CONCLUSIONS

Although conventional GA is a powerful tool in optimizing classifiers used in CAD schemes of medical images, it is very computationally intensive. This study demonstrated that using a new SFFS based approach enabled to significantly improve efficacy of image feature selection for developing CAD schemes.

A spatial shape constrained clustering method for mammographic mass segmentation

A novel clustering method is proposed for mammographic mass segmentation on extracted regions of interest (ROIs) by using deterministic annealing incorporating circular shape function (DACF). The objective function reported in this study uses both intensity and spatial shape information, and the dominant dissimilarity measure is controlled by two weighting parameters. As a result, pixels having similar intensity information but located in different regions can be differentiated. Experimental results shows that, by using DACF, the mass segmentation results in digitized mammograms are improved with optimal mass boundaries, less number of noisy patches, and computational efficiency. An average probability of segmentation error of 7.18% for well-defined masses (or 8.06% for ill-defined masses) was obtained by using DACF on MiniMIAS database, with 5.86% (or 5.55%) and 6.14% (or 5.27%) improvements as compared to the standard DA and fuzzy c-means methods.

Semi-automated segmentation and classification of digital breast tomosynthesis reconstructed images

Digital breast tomosynthesis (DBT) is a limited-angle tomographic x-ray imaging technique that reduces the effect of tissue superposition observed in planar mammography. An integrated imaging platform that combines DBT with near infrared spectroscopy (NIRS) to provide co-registered anatomical and functional imaging is under development. Incorporation of anatomic priors can benefit NIRS reconstruction. In this work, we provide a segmentation and classification method to extract potential lesions, as well as adipose, fibroglandular, muscle and skin tissue in reconstructed DBT images that serve as anatomic priors during NIRS reconstruction. The method may also be adaptable for estimating tumor volume, breast glandular content, and for extracting lesion features for potential application to computer aided detection and diagnosis.

Characterization of masses in digital breast tomosynthesis: comparison of machine learning in projection views and reconstructed slices

PURPOSE

In digital breast tomosynthesis (DBT), quasi-three-dimensional (3D) structural information is reconstructed from a small number of 2D projection view (PV) mammograms acquired over a limited angular range. The authors developed preliminary computer-aided diagnosis (CADx) methods for classification of malignant and benign masses and compared the effectiveness of analyzing lesion characteristics in the reconstructed DBT slices and in the PVs.

METHODS

A data set of MLO view DBT of 99 patients containing 107 masses (56 malignant and 51 benign) was collected at the Massachusetts General Hospital with IRB approval. The DBTs were obtained with a GE prototype system which acquired 11 PVs over a 50 degree arc. The authors reconstructed the DBTs at 1 mm slice interval using a simultaneous algebraic reconstruction technique. The region of interest (ROI) containing the mass was marked by a radiologist in the DBT volume and the corresponding ROIs on the PVs were derived based on the imaging geometry. The subsequent processes were fully automated. For classification of masses using the DBT-slice approach, the mass on each slice was segmented by an active contour model initialized with adaptive k-means clustering. A spiculation likelihood map was generated by analysis of the gradient directions around the mass margin and spiculation features were extracted from the map. The rubber band straightening transform (RBST) was applied to a band of pixels around the segmented mass boundary. The RBST image was enhanced by Sobel filtering in the horizontal and vertical directions, from which run-length statistics texture features were extracted. Morphological features including those from the normalized radial length were designed to describe the mass shape. A feature space composed of the spiculation features, texture features, and morphological features extracted from the central slice alone and seven feature spaces obtained by averaging the corresponding features from three to 19 slices centered at the central slice were compared. For classification of masses using the PV approach, a feature extraction process similar to that described above for the DBT approach was performed on the ROIs from the individual PVs. Six feature spaces obtained from the central PV alone and by averaging the corresponding features from three to 11 PVs were formed. In each feature space for either the DBT-slice or the PV approach, a linear discriminant analysis classifier with stepwise feature selection was trained and tested using a two-loop leave-one-case-out resampling procedure. Simplex optimization was used to guide feature selection automatically within the training set in each leave-one-case-out cycle. The performance of the classifiers was evaluated by the area (Az) under the receiver operating characteristic curve.

RESULTS

The test Az values from the DBT-slice approach ranged from 0.87 +/- 0.03 to 0.93 +/- 0.02, while those from the PV approach ranged from 0.78 +/- 0.04 to 0.84 +/- 0.04. The highest test Az of 0.93 +/- 0.02 from the nine-DBT-slice feature space was significantly (p = 0.006) better than the highest test Az of 0.84 +/- 0.04 from the nine-PV feature space.

CONCLUSION

The features of breast lesions extracted from the DBT slices consistently provided higher classification accuracy than those extracted from the PV images.

Discrimination of breast tumors in ultrasonic images using an ensemble classifier based on the AdaBoost algorithm with feature selection

This paper proposes a novel algorithm to estimate a log-compressed K distribution parameter and presents an algorithm to discriminate breast tumors in ultrasonic images. We computed a total of 208 features for discrimination, including those based on a parameter of a log-compressed K-distribution, which quantifies the homogeneity of the echo pattern in the tumor, but is influenced by compression parameters in the ultrasonic device. The proposed algorithm estimates the parameter of the log-compressed K-distribution in a manner free from this influence. To quantify irregularities in tumor shape, pattern-spectrum-based features were newly developed in this paper. The discrimination process uses an ensemble classifier trained by a multiclass AdaBoost learning algorithm (AdaBoost.M2), combined with a sequential feature-selection process. A 10-fold cross-validation test validated the performance, and the results were compared with those of a Mahalanobis distance-based classifier and a multiclass support vector machine. A total of 200 carcinomas, 50 fibroadenomas, and 50 cysts were used in the experiments. This paper demonstrates that the combination of a classifier trained by AdaBoost.M2 and features based on the estimated parameter of a log-compressed K-distribution, as well as those of the pattern spectrum, are useful for the discrimination of tumors.

A new automated method for the segmentation and characterization of breast masses on ultrasound images

Segmentation is one of the first steps in most computer-aided diagnosis systems for characterization of masses as malignant or benign. In this study, the authors designed an automated method for segmentation of breast masses on ultrasound (US) images. The method automatically estimated an initial contour based on a manually identified point approximately at the mass center. A two-stage active contour method iteratively refined the initial contour and performed self-examination and correction on the segmentation result. To evaluate the method, the authors compared it with manual segmentation by two experienced radiologists (R1 and R2) on a data set of 488 US images from 250 biopsy-proven masses (100 malignant and 150 benign). Two area overlap ratios (AOR1 and AOR2) and an area error measure were used as performance measures to evaluate the segmentation accuracy. Values for AOR1, defined as the ratio of the intersection of the computer and the reference segmented areas to the reference segmented area, were 0.82 +/- 0.16 and 0.84 +/- 0.18, respectively, when manually segmented mass regions by R1 and R2 were used as the reference. Although this indicated a high agreement between the computer and manual segmentations, the two radiologists' manual segmentation results were significantly (p < 0.03) more consistent, with AOR1 = 0.84 +/- 0.16 and 0.91 +/- 0.12, respectively, when the segmented regions by R1 and R2 were used as the reference. To evaluate the segmentation method in terms of lesion classification accuracy, feature spaces were formed by extracting texture, width-to-height, and posterior shadowing features based on either automated computer segmentation or the radiologists' manual segmentation. A linear discriminant analysis classifier was designed using stepwise feature selection and two-fold cross validation to characterize the mass as malignant or benign. For features extracted from computer segmentation, the case-based test A(z) values ranged from 0.88 +/- 0.03 to 0.92 +/- 0.02, indicating a comparable performance to those extracted from manual segmentation by radiologists (A(z) value range: 0.87 +/- 0.03 to 0.90 +/- 0.03).

Supervised pattern recognition for the prediction of contrast-enhancement appearance in brain tumors from multivariate magnetic resonance imaging and spectroscopy

OBJECTIVE

The purpose of this study was to develop a pattern classification algorithm for use in predicting the location of new contrast-enhancement in brain tumor patients using data obtained via multivariate magnetic resonance (MR) imaging from a prior scan. We also explore the use of feature selection or weighting in improving the accuracy of the pattern classifier.

METHODS AND MATERIALS

Contrast-enhanced MR images, perfusion images, diffusion images, and proton spectroscopic imaging data were obtained from 26 patients with glioblastoma multiforme brain tumors, divided into a design set and an unseen test set for verification of results. A k-NN algorithm was implemented to classify unknown data based on a set of training data with ground truth derived from post-treatment contrast-enhanced images; the quality of the k-NN results was evaluated using a leave-one-out cross-validation method. A genetic algorithm was implemented to select optimal features and feature weights for the k-NN algorithm. The binary representation of the weights was varied from 1 to 4 bits. Each individual parameter was thresholded as a simple classification technique, and the results compared with the k-NN.

RESULTS

The feature selection k-NN was able to achieve a sensitivity of 0.78+/-0.18 and specificity of 0.79+/-0.06 on the holdout test data using only 7 of the 38 original features. Similar results were obtained with non-binary weights, but using a larger number of features. Overfitting was also observed in the higher bit representations. The best single-variable classifier, based on a choline-to-NAA abnormality index computed from spectroscopic data, achieved a sensitivity of 0.79+/-0.20 and specificity of 0.71+/-0.11. The k-NN results had lower variation across patients than the single-variable classifiers.

CONCLUSIONS

We have demonstrated that the optimized k-NN rule could be used for quantitative analysis of multivariate images, and be applied to a specific clinical research question. Selecting features was found to be useful in improving the accuracy of feature weighting algorithms and improving the comprehensibility of the results. We believe that in addition to lending insight into parameter relevance, such algorithms may be useful in aiding radiological interpretation of complex multimodality datasets.

Characterization of mammographic masses based on level set segmentation with new image features and patient information

Computer-aided diagnosis (CAD) for characterization of mammographic masses as malignant or benign has the potential to assist radiologists in reducing the biopsy rate without increasing false negatives. The purpose of this study was to develop an automated method for mammographic mass segmentation and explore new image based features in combination with patient information in order to improve the performance of mass characterization. The authors' previous CAD system, which used the active contour segmentation, and morphological, textural, and spiculation features, has achieved promising results in mass characterization. The new CAD system is based on the level set method and includes two new types of image features related to the presence of microcalcifications with the mass and abruptness of the mass margin, and patient age. A linear discriminant analysis (LDA) classifier with stepwise feature selection was used to merge the extracted features into a classification score. The classification accuracy was evaluated using the area under the receiver operating characteristic curve. The authors' primary data set consisted of 427 biopsy-proven masses (200 malignant and 227 benign) in 909 regions of interest (ROIs) (451 malignant and 458 benign) from multiple mammographic views. Leave-one-case-out resampling was used for training and testing. The new CAD system based on the level set segmentation and the new mammographic feature space achieved a view-based Az value of 0.83 +/- 0.01. The improvement compared to the previous CAD system was statistically significant (p = 0.02). When patient age was included in the new CAD system, view-based and case-based Az values were 0.85 +/- 0.01 and 0.87 +/- 0.02, respectively. The study also demonstrated the consistency of the newly developed CAD system by evaluating the statistics of the weights of the LDA classifiers in leave-one-case-out classification. Finally, an independent test on the publicly available digital database for screening mammography with 132 benign and 197 malignant ROIs containing masses achieved a view-based Az value of 0.84 +/- 0.02.

Decision optimization of case-based computer-aided decision systems using genetic algorithms with application to mammography

This paper presents an optimization framework for improving case-based computer-aided decision (CB-CAD) systems. The underlying hypothesis of the study is that each example in the knowledge database of a medical decision support system has different importance in the decision making process. A new decision algorithm incorporating an importance weight for each example is proposed to account for these differences. The search for the best set of importance weights is defined as an optimization problem and a genetic algorithm is employed to solve it. The optimization process is tailored to maximize the system's performance according to clinically relevant evaluation criteria. The study was performed using a CAD system developed for the classification of regions of interests (ROIs) in mammograms as depicting masses or normal tissue. The system was constructed and evaluated using a dataset of ROIs extracted from the Digital Database for Screening Mammography (DDSM). Experimental results show that, according to receiver operator characteristic (ROC) analysis, the proposed method significantly improves the overall performance of the CAD system as well as its average specificity for high breast mass detection rates.

Computer-aided mass detection based on ipsilateral multiview mammograms

RATIONALE AND OBJECTIVES

Recent reports on advances in computer-aided detection (CAD) indicate that current schemes miss early-stage breast cancers and result in a relatively large false-positive detection rate in order to achieve a high sensitivity rate for mass detection. This paper is inspired by the interpretation procedure from mammographers. The abnormal diagnosis can be derived from multiple views but is not available through single-view image analysis.

MATERIALS AND METHODS

A new multiview CAD system for early-stage breast cancer detection, which is based on modifying the optimized CAD algorithms from our prior single-view CAD system for constructing an adaptive ipsilateral multiview concurrent CAD system, is presented in this paper. The selection and design for the training and testing ipsilateral multiview mammogram databases are described here.

RESULTS

The performance evaluation of the developed ipsilateral multiview CAD system using free-response receiver operating characteristic analysis and computerized receiver operating characteristic experiments are presented. The results indicated that the proposed multiview CAD system is significantly superior to the single-view CAD systems based on statistically standard P-values.

CONCLUSION

This paper addresses a very important and timely project. It is related to two main problems regarding the development of breast cancer detection and diagnosis: early-stage detection and diagnosis of breast cancer with digital mammogram, and overall improvement of CAD system performance for clinical implementation. In order to improve the efficacy, accuracy, and efficiency of the current CAD scheme, an entirely new class of CAD method is required. This paper is unique in that a comprehensive and state-of-the-art approach is proposed for the CAD scheme of digital mammography. From the design aspect of the CAD scheme, the proposed ipsilateral multiview CAD method is innovative and quite different from current single-view CAD methods.

False positive reduction in mammographic mass detection using local binary patterns

In this paper we propose a new approach for false positive reduction in the field of mammographic mass detection. The goal is to distinguish between the true recognized masses and the ones which actually are normal parenchyma. Our proposal is based on Local Binary Patterns (LBP) for representing salient micro-patterns and preserving at the same time the spatial structure of the masses. Once the descriptors are extracted, Support Vector Machines (SVM) are used for classifying the detected masses. We test our proposal using a set of 1792 suspicious regions of interest extracted from the DDSM database. Exhaustive experiments illustrate that LBP features are effective and efficient for false positive reduction even at different mass sizes, a critical aspect in mass detection systems. Moreover, we compare our proposal with current methods showing that LBP obtains better performance.

Computer-aided diagnosis of pulmonary nodules on CT scans: segmentation and classification using 3D active contours

We are developing a computer-aided diagnosis (CAD) system to classify malignant and benign lung nodules found on CT scans. A fully automated system was designed to segment the nodule from its surrounding structured background in a local volume of interest (VOI) and to extract image features for classification. Image segmentation was performed with a three-dimensional (3D) active contour (AC) method. A data set of 96 lung nodules (44 malignant, 52 benign) from 58 patients was used in this study. The 3D AC model is based on two-dimensional AC with the addition of three new energy components to take advantage of 3D information: (1) 3D gradient, which guides the active contour to seek the object surface, (2) 3D curvature, which imposes a smoothness constraint in the z direction, and (3) mask energy, which penalizes contours that grow beyond the pleura or thoracic wall. The search for the best energy weights in the 3D AC model was guided by a simplex optimization method. Morphological and gray-level features were extracted from the segmented nodule. The rubber band straightening transform (RBST) was applied to the shell of voxels surrounding the nodule. Texture features based on run-length statistics were extracted from the RBST image. A linear discriminant analysis classifier with stepwise feature selection was designed using a second simplex optimization to select the most effective features. Leave-one-case-out resampling was used to train and test the CAD system. The system achieved a test area under the receiver operating characteristic curve (A(z)) of 0.83 +/- 0.04. Our preliminary results indicate that use of the 3D AC model and the 3D texture features surrounding the nodule is a promising approach to the segmentation and classification of lung nodules with CAD. The segmentation performance of the 3D AC model trained with our data set was evaluated with 23 nodules available in the Lung Image Database Consortium (LIDC). The lung nodule volumes segmented by the 3D AC model for best classification were generally larger than those outlined by the LIDC radiologists using visual judgment of nodule boundaries.

A genetic algorithm design for microcalcification detection and classification in digital mammograms

In this paper, we propose a genetic algorithm design to automatically classify and detect micocalcification clusters in digital mammograms. The proposed GA technique is characterised by transforming input images into a feature domain, where each pixel is represented by its mean and standard deviation inside a surrounding window of size 9 x 9 pixel. In the feature domain, chromosomes are constructed to populate the initial generation and further features are extracted to enable the proposed GA to search for optimised classification and detection of microcalcification clusters via regions of 128 x 128 pixels. Extensive experiments show that the proposed GA design is able to achieve high performances in microcalcification classification and detection, which are measured by ROC curves, sensitivity against specificity, areas under ROC curves and benchmarked by existing representative techniques.

A KNOWLEDGE-BASED LUNG NODULE DETECTION SYSTEM FOR HELICAL CT IMAGES

In this paper, we describe a knowledge-based system for segmenting and labeling lung nodule on helical CT images. The system was developed under a blackboard environment that incorporates a lung knowledge model, image processing model, inference engine and a blackboard. Lung model, which contains both analogical and propositional knowledge about lung in the form of semantic networks, was used to guide the interpretation process. The system works in a hierarchical structure, from large structures to the final nodule candidates by focusing on the interested region step by step. The symbolic variables, introduced to accomplish high-level inference, were defined by fuzzy confidence functions in the lung model. Composite fuzzy functions were applied to evaluate the plausibility of the mapping between the image and lung model objects. Anatomical lung segments knowledge was embedded in the system to direct 3D validation of suspicious objects. Structures were identified and abnormal objects were reported. The experimental results obtained demonstrate the proof of concept and the potential of the automated knowledge-based lung nodule detection system.

Applying spatial distribution analysis techniques to classification of 3D medical images

OBJECTIVE

The objective of this paper is to classify 3D medical images by analyzing spatial distributions to model and characterize the arrangement of the regions of interest (ROIs) in 3D space.

METHODS AND MATERIAL

Two methods are proposed for facilitating such classification. The first method uses measures of similarity, such as the Mahalanobis distance and the Kullback-Leibler (KL) divergence, to compute the difference between spatial probability distributions of ROIs in an image of a new subject and each of the considered classes represented by historical data (e.g., normal versus disease class). A new subject is predicted to belong to the class corresponding to the most similar dataset. The second method employs the maximum likelihood (ML) principle to predict the class that most likely produced the dataset of the new subject.

RESULTS

The proposed methods have been experimentally evaluated on three datasets: synthetic data (mixtures of Gaussian distributions), realistic lesion-deficit data (generated by a simulator conforming to a clinical study), and functional MRI activation data obtained from a study designed to explore neuroanatomical correlates of semantic processing in Alzheimer's disease (AD).

CONCLUSION

Performed experiments demonstrated that the approaches based on the KL divergence and the ML method provide superior accuracy compared to the Mahalanobis distance. The later technique could still be a method of choice when the distributions differ significantly, since it is faster and less complex. The obtained classification accuracy with errors smaller than 1% supports that useful diagnosis assistance could be achieved assuming sufficiently informative historic data and sufficient information on the new subject.

Standardization for image characteristics in telemammography using genetic and nonlinear algorithms

As the soft copy reading and computer assisted diagnosis (CAD) in mammography become more and more important, the standardization of digital images becomes paramount. Telemammography and telemedicine requires the standardization for image characteristics, such as image resolution, bit-depth and intensity response. Soft copy reading and CAD in mammography are both dependent on the characteristics of the source of the digital data, either direct digital mammography or digitized screen-film mammography. An algorithm developed on images from one database may not perform well as on images from another database (with a different digitization). In this paper, we describe two methods based on a genetic algorithm and a nonlinear algorithm for standardization of digitized and digital mammography. The proposed standardization techniques are based on geometric and intensity transformations that are discovered using a set of calibration images. A set of transformation algorithm is used to search for the best standardization.

Steepest changes of a probability-based cost function for delineation of mammographic masses: A validation study

Our purpose in this work was to develop an automatic boundary detection method for mammographic masses and to rigorously test this method via statistical analysis. The segmentation method utilized a steepest change analysis technique for determining the mass boundaries based on a composed probability density cost function. Previous investigators have shown that this function can be utilized to determine the border of the mass body. We have further analyzed this method and have discovered that the steepest changes in this function can produce mass delineations that include extended projections. The method was tested on 124 digitized mammograms selected from the University of South Florida's Digital Database for Screening Mammography (DDSM). The segmentation results were validated using overlap, accuracy, sensitivity, and specificity statistics, where the gold standards were manual traces provided by two expert radiologists. We have concluded that the best intensity threshold corresponds to a particular steepest change location within the composed probability density function. We also found that our results are more closely correlated with one expert than with the second expert. These findings were verified via Analysis of Variance (ANOVA) testing. The ANOVA tests obtained p-values ranging from 1.03 x 10(-2)-7.51 x 10(-17) for the single observer studies and 2.03 x 10(-2)-9.43 x 10(-4) for the two observer studies. Results were categorized using three significance levels, i.e., p<0.001 (extremely significant), p <0.01 (very significant), and p <0.05 (significant), respectively.

Incorporation of an iterative, linear segmentation routine into a mammographic mass CAD system

In previous research, we have developed a computer-aided detection (CAD) system designed to detect masses in mammograms. The previous version of our system employed a simple but imprecise method to localize the masses. In this research, we present a more robust segmentation routine for use with mammographic masses. Our hypothesis is that by more accurately describing the morphology of the masses, we can improve the CAD system's ability to distinguish masses from other mammographic structures. To test this hypothesis, we incorporated the new segmentation routine into our CAD system and examined the change in performance. The developed iterative, linear segmentation routine is a gray level-based procedure. Using the identified regions from the previous CAD system as the initial seeds, the new segmentation algorithm refines the suspicious mass borders by making estimates of the interior and exterior pixels. These estimates are then passed to a linear discriminant, which determines the optimal threshold between the interior and exterior pixels. After applying the threshold and identifying the object's outline, two constraints on the border are applied to reduce the influence of background noise. After the border is constrained, the process repeats until a stopping criterion is reached. The segmentation routine was tested on a study database of 183 mammographic images extracted from the Digital Database for Screening Mammography. Eighty-three of the images contained 50 malignant and 50 benign masses; 100 images contained no masses. The previously developed CAD system was used to locate a set of suspicious regions of interest (ROIs) within the images. To assess the performance of the segmentation algorithm, a set of 20 features was measured from the suspicious regions before and after the application of the developed segmentation routine. Receiver operating characteristic (ROC) analysis was employed on the ROIs to examine the discriminatory capabilities of each individual feature before and after the segmentation routine. A statistically significant performance increase was found in many of the individual features, particularly those describing the mass borders. To examine how the incorporation of the segmentation routine affected the performance of the overall CAD system, free-response ROC (FROC) analysis was employed. When considering only malignant masses, the FROC performance of the system with the segmentation routine appeared better than the previous system. When detecting 90% of the malignant masses, the previous system achieved 4.9 false positives per image (FPpI) compared to the post-segmentation system's 4.2 FPpI. At 80% sensitivity, the respective FPpI were 3.5 and 1.6.

Lung nodule detection on thoracic computed tomography images: preliminary evaluation of a computer-aided diagnosis system

We are developing a computer-aided diagnosis (CAD) system for lung nodule detection on thoracic helical computed tomography (CT) images. In the first stage of this CAD system, lung regions are identified by a k-means clustering technique. Each lung slice is classified as belonging to the upper, middle, or the lower part of the lung volume. Within each lung region, structures are segmented again using weighted k-means clustering. These structures may include true lung nodules and normal structures consisting mainly of blood vessels. Rule-based classifiers are designed to distinguish nodules and normal structures using 2D and 3D features. After rule-based classification, linear discriminant analysis (LDA) is used to further reduce the number of false positive (FP) objects. We performed a preliminary study using 1454 CT slices from 34 patients with 63 lung nodules. When only LDA classification was applied to the segmented objects, the sensitivity was 84% (53/63) with 5.48 (7961/1454) FP objects per slice. When rule-based classification was used before LDA, the free response receiver operating characteristic (FROC) curve improved over the entire sensitivity and specificity ranges of interest. In particular, the FP rate decreased to 1.74 (2530/1454) objects per slice at the same sensitivity. Thus, compared to FP reduction with LDA alone, the inclusion of rule-based classification lead to an improvement in detection accuracy for the CAD system. These preliminary results demonstrate the feasibility of our approach to lung nodule detection and FP reduction on CT images.

Development and evaluation of a case-based reasoning classifier for prediction of breast biopsy outcome with BI-RADS lexicon

Approximately 70-85% of breast biopsies are performed on benign lesions. To reduce this high number of biopsies performed on benign lesions, a case-based reasoning (CBR) classifier was developed to predict biopsy results from BI-RADS findings. We used 1433 (931 benign) biopsy-proven mammographic cases. CBR similarity was defined using either the Hamming or Euclidean distance measure over case features. Ten features represented each case: calcification distribution, calcification morphology, calcification number, mass margin, mass shape, mass density, mass size, associated findings, special cases, and age. Performance was evaluated using Round Robin sampling, Receiver Operating Characteristic (ROC) analysis, and bootstrap. To determine the most influential features for the CBR, an exhaustive feature search was performed over all possible feature combinations (1022) and similarity thresholds. Influential features were defined as the most frequently occurring features in the feature subsets with the highest partial ROC areas (0.90AUC). For CBR with Hamming distance, the most influential features were found to be mass margin, calcification morphology, age, calcification distribution, calcification number, and mass shape, resulting in an 0.90AUC of 0.33. At 95% sensitivity, the Hamming CBR would spare from biopsy 34% of the benign lesions. At 98% sensitivity, the Hamming CBR would spare 27% benign lesions. For the CBR with Euclidean distance, the most influential feature subset consisted of mass margin, calcification morphology, age, mass density, and associated findings, resulting in 0.90AUC of 0.37. At 95% sensitivity, the Euclidean CBR would spare from biopsy 41% benign lesions. At 98% sensitivity, the Euclidean CBR would spare 27% benign lesions. The profile of cases spared by both distance measures at 98% sensitivity indicates that the CBR is a potentially useful diagnostic tool for the classification of mammographic lesions, by recommending short-term follow-up for likely benign lesions that is in agreement with final biopsy results and mammographer's intuition.

Application of the mutual information criterion for feature selection in computer-aided diagnosis

The purpose of this study was to investigate an information theoretic approach to feature selection for computer-aided diagnosis (CAD). The approach is based on the mutual information (MI) concept. MI measures the general dependence of random variables without making any assumptions about the nature of their underlying relationships. Consequently, MI can potentially offer some advantages over feature selection techniques that focus only on the linear relationships of variables. This study was based on a database of statistical texture features extracted from perfusion lung scans. The ultimate goal was to select the optimal subset of features for the computer-aided diagnosis of acute pulmonary embolism (PE). Initially, the study addressed issues regarding the approximation of MI in a limited dataset as it is often the case in CAD applications. The MI selected features were compared to those features selected using stepwise linear discriminant analysis and genetic algorithms for the same PE database. Linear and nonlinear decision models were implemented to merge the selected features into a final diagnosis. Results showed that the MI is an effective feature selection criterion for nonlinear CAD models overcoming some of the well-known limitations and computational complexities of other popular feature selection techniques in the field.

Computer-aided characterization of mammographic masses: accuracy of mass segmentation and its effects on characterization

Mass segmentation is used as the first step in many computer-aided diagnosis (CAD) systems for classification of breast masses as malignant or benign. The goal of this paper was to study the accuracy of an automated mass segmentation method developed in our laboratory, and to investigate the effect of the segmentation stage on the overall classification accuracy. The automated segmentation method was quantitatively compared with manual segmentation by two expert radiologists (R1 and R2) using three similarity or distance measures on a data set of 100 masses. The area overlap measures between R1 and R2, the computer and R1, and the computer and R2 were 0.76 +/- 0.13, 0.74 +/- 0.11, and 0.74 +/- 0.13, respectively. The interobserver difference in these measures between the two radiologists was compared with the corresponding differences between the computer and the radiologists. Using three similarity measures and data from two radiologists, a total of six statistical tests were performed. The difference between the computer and the radiologist segmentation was significantly larger than the interobserver variability in only one test. Two sets of texture, morphological, and spiculation features, one based on the computer segmentation, and the other based on radiologist segmentation, were extracted from a data set of 249 films from 102 patients. A classifier based on stepwise feature selection and linear discriminant analysis was trained and tested using the two feature sets. The leave-one-case-out method was used for data sampling. For case-based classification, the area Az under the receiver operating characteristic (ROC) curve was 0.89 and 0.88 for the feature sets based on the radiologist segmentation and computer segmentation, respectively. The difference between the two ROC curves was not statistically significant.

Performance gain in computer-assisted detection schemes by averaging scores generated from artificial neural networks with adaptive filtering

The authors investigated a new method to optimize artificial neural networks (ANNs) with adaptive filtering used in computer-assisted detection schemes in digitized mammograms and to assess performance changes when averaging classification scores from three sets of optimized schemes. Two independent training and testing image databases involving 978 and 830 digitized mammograms, respectively, were used in this study. In the training data set, initial filtering and subtraction resulted in the identification of 592 mass regions and 3790 suspicious, but actually negative regions. These regions (including both true-positive and negative regions) were segmented into three subsets three times based on the calculation of the values of three features as segmentation indices. The indices were "mass" size multiplied by their digital value contrast, conspicuity, and circularity. Nine ANN-based classifiers were separately optimized using a genetic algorithm for each subset of regions. Each region was assigned three classification scores after applying the three adaptive ANNs. The performance gain of the CAD scheme after averaging the three scores for each suspicious region was tested using an independent data set and a ROC methodology. The experimental results showed that the areas under ROC curves (Az) for the testing database using three sets of optimized ANNs individually were 0.84+/-0.01, 0.83+/-0.01, and 0.84+/-0.01, respectively. The between-index correlations of three A values were 0.013, -0.007, and 0.086. Similar to averaging diagnostic ratings from independent observers, by averaging three ANN-generated scores for each testing region, the performance of the CAD scheme was significantly improved (p<0.001) with Az value of 0.95+/-0.01.

Improvement of mammographic mass characterization using spiculation meausures and morphological features

We are developing new computer vision techniques for characterization of breast masses on mammograms. We had previously developed a characterization method based on texture features. The goal of the present work was to improve our characterization method by making use of morphological features. Toward this goal, we have developed a fully automated, three-stage segmentation method that includes clustering, active contour, and spiculation detection stages. After segmentation, morphological features describing the shape of the mass were extracted. Texture features were also extracted from a band of pixels surrounding the mass. Stepwise feature selection and linear discriminant analysis were employed in the morphological, texture, and combined feature spaces for classifier design. The classification accuracy was evaluated using the area Az under the receiver operating characteristic curve. A data set containing 249 films from 102 patients was used. When the leave-one-case-out method was applied to partition the data set into trainers and testers, the average test Az for the task of classifying the mass on a single mammographic view was 0.83 +/- 0.02, 0.84 +/- 0.02, and 0.87 +/- 0.02 in the morphological, texture, and combined feature spaces, respectively. The improvement obtained by supplementing texture features with morphological features in classification was statistically significant (p = 0.04). For classifying a mass as malignant or benign, we combined the leave-one-case-out discriminant scores from different views of a mass to obtain a summary score. In this task, the test Az value using the combined feature space was 0.91 +/- 0.02. Our results indicate that combining texture features with morphological features extracted from automatically segmented mass boundaries will be an effective approach for computer-aided characterization of mammographic masses.

False-positive reduction in CAD mass detection using a competitive classification strategy

High false-positive (FP) rate remains to be one of the major problems to be solved in CAD study because too many false-positively cued signals will potentially degrade the performance of detecting true-positive regions and increase the call-back rate in CAD environment. In this paper, we proposed a novel classification method for FP reduction, where the conventional "hard" decision classifier is cascaded with a "soft" decision classification with the objective to reduce false-positives in the cases with multiple FPs retained after the "hard" decision classification. The "soft" classification takes a competitive classification strategy in which only the "best" ones are selected from the pre-classified suspicious regions as the true mass in each case. A neural network structure is designed to implement the proposed competitive classification. Comparative studies of FP reduction on a database of 79 images by a "hard" decision classification and a combined "hard"-"soft" classification method demonstrated the efficiency of the proposed classification strategy. For example, for the high FP sub-database which has only 31.7% of total images but accounts for 63.5% of whole FPs generated in single "hard" classification, the FPs can be reduced for 56% (from 8.36 to 3.72 per image) by using the proposed method at the cost of 1% TP loss (from 69% to 68%) in whole database, while it can only be reduced for 27% (from 8.36 to 6.08 per image) by simply increasing the threshold of "hard" classifier with a cost of TP loss as high as 14% (from 69% to 55%). On the average in whole database, the FP reduction by hybrid "hard"-"soft" classification is 1.58 per image as compared to 1.11 by "hard" classification at the TP costs described above. Because the cases with high dense tissue are of higher risk of cancer incidence and false-negative detection in mammogram screening, and usually generate more FPs in CAD detection, the method proposed in this paper will be very helpful in improving the performance of early detection of breast cancer with CAD.

Evolutionary computation in medicine: an overview

The term evolutionary computation encompasses a host of methodologies inspired by natural evolution that are used to solve hard problems. This paper provides an overview of evolutionary computation as applied to problems in the medical domains. We begin by outlining the basic workings of six types of evolutionary algorithms: genetic algorithms, genetic programming, evolution strategies, evolutionary programming, classifier systems, and hybrid systems. We then describe how evolutionary algorithms are applied to solve medical problems, including diagnosis, prognosis, imaging, signal processing, planning, and scheduling. Finally, we provide an extensive bibliography, classified both according to the medical task addressed and according to the evolutionary technique used.

Segmentation of suspicious clustered microcalcifications in mammograms

We have developed a multistage computer-aided diagnosis (CAD) scheme for the automated segmentation of suspicious microcalcification clusters in digital mammograms. The scheme consisted of three main processing steps. First, the breast region was segmented and its high-frequency content was enhanced using unsharp masking. In the second step, individual microcalcifications were segmented using local histogram analysis on overlapping subimages. For this step, eight histogram features were extracted for each subimage and were used as input to a fuzzy rule-based classifier that identified subimages containing microcalcifications and assigned the appropriate thresholds to segment any microcalcifications within them. The final step clustered the segmented microcalcifications and extracted the following features for each cluster: the number of microcalcifications, the average distance between microcalcifications, and the average number of times pixels in the cluster were segmented in the second step. Fuzzy logic rules incorporating the cluster features were designed to remove nonsuspicious clusters, defined as those with typically benign characteristics. A database of 98 images, with 48 images containing one or more microcalcification clusters, provided training and testing sets to optimize the parameters and evaluate the CAD scheme, respectively. The results showed a true positive rate of 93.2% and an average of 0.73 false positive clusters per image. A comparison of our results with other reported segmentation results on the same database showed comparable sensitivity and at the same time an improved false positive rate. The performance of the CAD scheme is encouraging for its use as an automatic tool for efficient and accurate diagnosis of breast cancer.

Pulmonary CT image classification with evolutionary programming

RATIONALE AND OBJECTIVES

It is often difficult to classify information in medical images from derived features. The purpose of this research was to investigate the use of evolutionary programming as a tool for selecting important features and generating algorithms to classify computed tomographic (CT) images of the lung.

MATERIALS AND METHODS

Training and test sets consisting of 11 features derived from multiple lung CT images were generated, along with an indicator of the target area from which features originated. The images included five parameters based on histogram analysis, 11 parameters based on run length and co-occurrence matrix measures, and the fractal dimension. Two classification experiments were performed. In the first, the classification task was to distinguish between the subtle but known differences between anterior and posterior portions of transverse lung CT sections. The second classification task was to distinguish normal lung CT images from emphysematous images. The performance of the evolutionary programming approach was compared with that of three statistical classifiers that used the same training and test sets.

RESULTS

Evolutionary programming produced solutions that compared favorably with those of the statistical classifiers. In separating the anterior from the posterior lung sections, the evolutionary programming results were better than two of the three statistical approaches. The evolutionary programming approach correctly identified all the normal and abnormal lung images and accomplished this by using less features than the best statistical method.

CONCLUSION

The results of this study demonstrate the utility of evolutionary programming as a tool for developing classification algorithms.

Classification of malignant and benign masses based on hybrid ART2LDA approach

A new type of classifier combining an unsupervised and a supervised model was designed and applied to classification of malignant and benign masses on mammograms. The unsupervised model was based on an adaptive resonance theory (ART2) network which clustered the masses into a number of separate classes. The classes were divided into two types: one containing only malignant masses and the other containing a mix of malignant and benign masses. The masses from the malignant classes were classified by ART2. The masses from the mixed classes were input to a supervised linear discriminant classifier (LDA). In this way, some malignant masses were separated and classified by ART2 and the less distinguishable benign and malignant masses were classified by LDA. For the evaluation of classifier performance, 348 regions of interest (ROI's) containing biopsy proven masses (169 benign and 179 malignant) were used. Ten different partitions of training and test groups were randomly generated using an average of 73% of ROI's for training and 27% for testing. Classifier design, including feature selection and weight optimization, was performed with the training group. The test group was kept independent of the training group. The performance of the hybrid classifier was compared to that of an LDA classifier alone and a backpropagation neural network (BPN). Receiver operating characteristics (ROC) analysis was used to evaluate the accuracy of the classifiers. The average area under the ROC curve (A(z)) for the hybrid classifier was 0.81 as compared to 0.78 for the LDA and 0.80 for the BPN. The partial areas above a true positive fraction of 0.9 were 0.34, 0.27 and 0.31 for the hybrid, the LDA and the BPN classifier, respectively. These results indicate that the hybrid classifier is a promising approach for improving the accuracy of classification in CAD applications.

The effects of co-occurrence matrix based texture parameters on the classification of solitary pulmonary nodules imaged on computed tomography

In this project, patients with a solitary pulmonary nodule, were imaged using high resolution computed tomography. Quantitative measures of texture were extracted from these images using co-occurrence matrices. These matrices were formed with different combinations of gray level quantization, distance between pixels and angles. The derived measures were input to a linear discriminant classifier to predict the classification (benign or malignant) of each nodule. Using a relative quantization scheme with eight levels, four features yielded an area under the ROC curve (Az) of 0.992; 93.8% (30/32) of cases were correctly classified when training and testing on the same cases; while 90.6% (29/32) were correctly classified when jackknifing was used.

Feature selection with limited datasets

Computer-aided diagnosis has the potential of increasing diagnostic accuracy by providing a second reading to radiologists. In many computerized schemes, numerous features can be extracted to describe suspect image regions. A subset of these features is then employed in a data classifier to determine whether the suspect region is abnormal or normal. Different subsets of features will, in general, result in different classification performances. A feature selection method is often used to determine an "optimal" subset of features to use with a particular classifier. A classifier performance measure (such as the area under the receiver operating characteristic curve) must be incorporated into this feature selection process. With limited datasets, however, there is a distribution in the classifier performance measure for a given classifier and subset of features. In this paper, we investigate the variation in the selected subset of "optimal" features as compared with the true optimal subset of features caused by this distribution of classifier performance. We consider examples in which the probability that the optimal subset of features is selected can be analytically computed. We show the dependence of this probability on the dataset sample size, the total number of features from which to select, the number of features selected, and the performance of the true optimal subset. Once a subset of features has been selected, the parameters of the data classifier must be determined. We show that, with limited datasets and/or a large number of features from which to choose, bias is introduced if the classifier parameters are determined using the same data that were employed to select the "optimal" subset of features.

A neural network to predict symptomatic lung injury

A nonlinear neural network that simultaneously uses pre-radiotherapy (RT) biological and physical data was developed to predict symptomatic lung injury. The input data were pre-RT pulmonary function, three-dimensional treatment plan doses and demographics. The output was a single value between 0 (asymptomatic) and 1 (symptomatic) to predict the likelihood that a particular patient would become symptomatic. The network was trained on data from 97 patients for 400 iterations with the goal to minimize the mean-squared error. Statistical analysis was performed on the resulting network to determine the model's accuracy. Results from the neural network were compared with those given by traditional linear discriminate analysis and the dose-volume histogram reduction (DVHR) scheme of Kutcher. Receiver-operator characteristic (ROC) analysis was performed on the resulting network which had Az = 0.833 +/- 0.04. (Az is the area under the ROC curve.) Linear discriminate multivariate analysis yielded an Az = 0.813 +/- 0.06. The DVHR method had Az = 0.521 +/- 0.08. The network was also used to rank the significance of the input variables. Future studies will be conducted to improve network accuracy and to include functional imaging data.

Multiobjective genetic optimization of diagnostic classifiers with implications for generating receiver operating characteristic curves

It is well understood that binary classifiers have two implicit objective functions (sensitivity and specificity) describing their performance. Traditional methods of classifier training attempt to combine these two objective functions (or two analogous class performance measures) into one so that conventional scalar optimization techniques can be utilized. This involves incorporating a priori information into the aggregation method so that the resulting performance of the classifier is satisfactory for the task at hand. We have investigated the use of a niched Pareto multiobjective genetic algorithm (GA) for classifier optimization. With niched Pareto GA's, an objective vector is optimized instead of a scalar function, eliminating the need to aggregate classification objective functions. The niched Pareto GA returns a set of optimal solutions that are equivalent in the absence of any information regarding the preferences of the objectives. The a priori knowledge that was used for aggregating the objective functions in conventional classifier training can instead be applied post-optimization to select from one of the series of solutions returned from the multiobjective genetic optimization. We have applied this technique to train a linear classifier and an artificial neural network (ANN), using simulated datasets. The performances of the solutions returned from the multiobjective genetic optimization represent a series of optimal (sensitivity, specificity) pairs, which can be thought of as operating points on a receiver operating characteristic (ROC) curve. All possible ROC curves for a given dataset and classifier are less than or equal to the ROC curve generated by the niched Pareto genetic optimization.

Feature selection for computerized mass detection in digitized mammograms by using a genetic algorithm

RATIONALE AND OBJECTIVES

To investigate optimization of feature selection for computerized mass detection in digitized mammograms, and to compare the effectiveness of a genetic algorithm (GA) in such optimization with that of an "exhaustive" search of all feature permutations.

MATERIALS AND METHODS

A Bayesian belief network (BBN) was used to classify positive and negative regions for masses depicted in digitized mammograms; 20 features were computed for each of 592 positive and 3,790 negative regions in two databases. Conditional probabilities for the BBN were computed by using a "training" database of 288 positive and 2,204 negative regions. Performance was measured by the area under the receiver operating characteristic curve (A) by using the remainder database (304 positive and 1,586 negative regions). The optimal set was first found by using an "exhaustive" (complete permutation) searching method. A GA-based search for the optimal set then was applied, and the results of the two approaches were compared.

RESULTS

As the number of features in the classifier increased, the A value increased until it reached a maximum performance for 11 features of 0.876 +/- 0.008. The A value then decreased monotonically as the number of features increased from 11 to 20. Using 100 random chromosomes (seeds) in the first generation, the GA identified the same optimal set of features but reduced the total computation time by a factor of 65.

CONCLUSION

A GA-based search might be an efficient and effective approach to selecting an optimal feature set.